CN106443664A - Radar and ESM track correlation method based on topology information under system error - Google Patents

Abstract

The invention discloses a radar and ESM (Electronic Support Measure) track correlation algorithm based on topology information under system error. For radar and ESM configured in different places, when the bearing lines intersect, a plurality of cross location points are produced, including wrong ghost points and correct correlated points. After all cross location points are obtained, partial correlation is performed first based on distance information, and a lot of impossible correlations are ruled out, so that the amount of calculation is reduced, and time is saved. Then, the remaining cross location points are divided and combined, the degree of correlation between each combination and radar measurement is calculated based on topology information, the cross location points in the combination corresponding to the highest degree of correlation are sequentially correlated with corresponding radar measurement, and thus, the correlation between all radar and ESM targets is obtained. Compared with the traditional track correlation algorithm, the topology information between targets is fully utilized, the influence of system error on track correlation can be well restrained, the rate of correct association is improved, and the rate of error correlation and the rate of missing correlation are reduced.

Description

Radar based on topology information and ESM Data Association under systematic error

Technical field

The invention belongs to Dissimilar sensors track association field, provide under a kind of systematic error existence condition based on topology The radar of information and electronic support measure (Electronic Support Measurement, ESM) Data Association.

Background technology

The information that same type of sensor provides is single, and is easily disturbed.Increasingly complicated with battlefield surroundings, Dissimilar sensors system System plays more and more important effect.Radar and ESM are two kinds of typical, most widely used active and passive sensors. Radar can provide accurate positional information, and ESM is provided that accurate attribute information, and both combine can more accurately, completely Ground understanding target, can solve the problems, such as target be who and target where, this is to carrying out Target threat estima tion, early warning and enter The operational commanding of one step and decision-making have vital meaning.

Radar and ESM track association be radar with the premise of ESM Track Fusion with crucial.Traditional radar and ESM flight path Correlating method usually assumes that error is the white Gaussian noise of zero-mean, by constructing association statisticses, then uses statistical theory side Method goes to adjudicate, and often have ignored the presence of systematic error.Because systematic error is unknown, association cannot be known with statistical method The thresholding of statistic, so can lead to produce substantial amounts of erroneous association during practical application associate it is impossible to meet demand with leakage.Right In the situation of strange land configuration, radar and ESM rhumb line phase intersection produce some cross bearing points, how to go to solve using these information Certainly the radar under systematic error and ESM flight path robust related question are the study hotspots in this field current.

Content of the invention

Radar for strange land configuration and ESM, when there is not error, the intersection of all radar measurements and associated is fixed The position of position should be overlapped；When only existing random meausrement error although incomplete overlap, but in the position in space it is Very close；When there is systematic error, the position of cross bearing point and radar measurement there occurs rotation and translate.But from See on the whole, systematic error has no effect on the relative position between target, i.e. the intersection of all radar measurements and all associateds Topological structure between anchor point is not affected by systematic error, and both pass through rotation and translation can approximately overlap.

In order to solve the above problems, the invention provides a kind of radar based on topology information and ESM Data Association. For strange land configuration radar and ESM, rhumb line phase intersection produce some cross bearing points, wherein comprise mistake ghost point and just The really point of association.After obtaining all cross bearing points, the present invention first carries out local association based on range information, and exclusion is a large amount of not Situation about may associate.Then to remaining cross bearing dot-dash subassembly, based on topology information calculate respectively various combination with The correlation degree of radar measurement, the corresponding cross bearing point of the maximum combination of correlation degree is closed with corresponding radar measurement successively Connection, thus obtain the incidence relation of all radars and ESM target.

The technical solution adopted for the present invention to solve the technical problems is：Thunder based on topology information under a kind of systematic error Reach and ESM Data Association, comprise the steps：

Step one, acquisition cross bearing point；

Step 2, the local association based on range information；

Step 3, to remaining cross bearing dot-dash subassembly；

Step 4, the overall gray relative based on topology information.

Compared to existing radar compared with ESM Data Association, the positive effect of the present invention is：Make full use of radar Measure the topology information with cross bearing point, can be good at the impact that suppression system error causes to track association, just improve Really association rate, reduces association rate and leakage association rate by mistake；By the local association based on range information, significantly reduce calculating Amount, has saved operation time it is ensured that the real-time of method.

Brief description

Fig. 1 is the radar based on topology information and ESM Data Association flow chart under systematic error；

Fig. 2 is cross bearing point schematic diagram when there is three targets.

Specific embodiment

Under a kind of systematic error proposed by the present invention, the radar based on topology information and ESM Data Association flow process are as schemed Shown in 1.

Assume that the radar of two strange land configurations and ESM detect to target under two-dimensional Cartesian system simultaneously, its Middle radar is located at coordinate (x_A,y_A) place, ESM is positioned at (x_B,y_B) place.If having n target, radar and ESM in k moment search coverage Measurement affected by random meausrement error and systematic error simultaneously, the angle measuring system error of radar range finding, angle measurement and ESM is divided Wei not constant △_r、△_θAAnd △_θB.Radar angle measurement, range finding and ESM angle measurement random error ε_θA(k)、ε_r(k) and ε_θBK () obeys all It is worth for 0, variance is respectivelyWithGaussian Profile.

Referring to the drawings 1, the radar based on topology information and ESM Data Association under a kind of systematic error, including as follows Several steps：

Step one, acquisition cross bearing point

Each radar target is sorted according to the incremental order of angle, obtains k moment radar target measurement sequence Z (k)={ Z_i (k) | i=1,2 ..., n }, wherein

It is the measurement of k i-th radar target of moment, r_j(k) andIt is the distance of k i-th radar target of moment respectively With azimuthal measurement value,WithThe true value in target range and orientation respectively.Carry out rectangular co-ordinate conversion, obtain conversion and sit Mark sequence Y (k)={ Y_i(k) | i=1,2 ..., n }, wherein Y_i(k)=[x_i(k) y_i(k)]^TIt is i-th radar target of k moment Converted measurement, x_i(k), y_iK () is respectively the transverse and longitudinal axial coordinate after changing.Equally, k moment ESM measurement point is passed according to angle Increasing order is arranged, and is designated as

WhereinFor the angle measurement of k j-th target of moment ESM,The angle true value of target.With radar it is Fusion center, obtains cross bearing coordinates matrixWithElement Ω^x(i, j) and Ω^y(i, j) records i-th thunder respectively Reach the angular cross anchor point x of target and j-th ESM target_ijAnd y_ijCoordinate：

Fig. 2 is the schematic diagram of cross bearing point when there is three targets.

Step 2, the local association based on range information

The angular cross anchor point of i-th radar target and j-th ESM target can be obtained away from radar apart from d_ij(k)

d_ijK the variance of () is

Make D_ij(k)=d_ij(k)-r_iK (), because the measure error between radar angular and distance and the angle of ESM is mutual Independent, so D_ijK the variance of () is

Local based on distance thick association process rule is：WhenWhen, that is, think i-th radar Target is derived from same target with j-th ESM target, i.e. two flight path test associations；When being unsatisfactory for it is believed that i-th radar mesh Mark and j-th ESM target can not possibly be derived from same target.

Step 3, to remaining cross bearing dot-dash subassembly

Define an ESM and radar potential track incidence matrix A_n×n, the i-th row jth column element a of A_ijRepresent and associate through thick The situation of i-th radar and j-th ESM track association, a after process_ijIt is Boolean type variable, a_ij=1 represents that satisfaction is thick closes bracing Part, that is, i-th radar may associate with j-th ESM flight path；a_ijThe thick Correlation Criteria of=0 foot with thumb down, that is, i-th radar with J-th ESM flight path can not possibly associate.

Because track association matrix A reflects that the possible of all ESM and radar target associates situation, when obtaining ESM and radar Combine it is possible to list all N kinds ESM with the possibility that radar track associates after potential track incidence matrix A, may by each Combination is described as a track association occurrence matrix B_l,

Wherein, l=1,2 ..., N,For track association occurrence matrix B_lI-th row jth column element, also for Boolean type variable.

According to the principle with an ESM target association for each radar target, can be to ESM and radar track incidence matrix A is split：From every a line of track association matrix, select unique 1 element, as track association occurrence matrix at this Row and the unique nonzero element of this row.After having split ESM and radar track incidence matrix matrix, because it is the relevant situation of institute Summation, necessarily comprise correctly to associate situation, ensuing work be exactly find from all track association occurrence matrixs Similar association.

Step 4, the overall gray relative based on topology information

Figure it is seen that systematic error has no effect on the relative position (i.e. the topological structure of object space) between target. The space topological of target is described as the alternate position spike sequence vector of all neighbours and this target by the present invention, and neighbours still pass according to angle The order arrangement increasing, using overall grey correlation analysis from the above-mentioned N number of track association occurrence matrix obtaining, to find and radar Most like one of object space topological structure.

Ensuing process is both for k time data, in order to describe conveniently, omits time variable k.I-th radar of k moment The topology information of point mark is the alternate position spike sequence vector of itself and all neighbours, As reference sequence, wherein i=1,2 ... n, V_i ¹And V_i ¹It is respectively the transverse and longitudinal axial coordinate of V.Splitting potential track incidence matrix can Obtain N number of track association occurrence matrix corresponding cross bearing point coordinates sequence, be defined as W_l={ W_l(i) | i=1,2 ... n }, (l=1,2 ... N), wherein W_lI () is the corresponding coordinate of the unique nonzero element of the i-th row of l-th track association occurrence matrix, can Know

Wherein,W respectively_lThe transverse and longitudinal axial coordinate of (i),For track association occurrence matrix B_lI-th row jth Column element.Then the space topological of l-th track association i-th target of occurrence matrix is described as sequence

I-th target of l-th track association occurrence matrix with the space topological coefficient correlation sequence of i-th radar plot be

I=1,2 ... n, l=1,2 ..., N, j=1,2 ..., n-1, m=1,2.NoteThen

In formula：ρ is resolution ratio, and ρ is less, and resolving power is poorer, generally takes ρ=0.5.

Because l-th track association occurrence matrix has n crosspoint, each crosspoint comprises n-1 neighbours, and each Cross bearing point coordinates sequence all comprises two coefficient correlation phasesequence component of transverse and longitudinal axle, defines l-th track association occurrence matrix Gray relation grades with the similarity of the space topological of radar target are：

It is calculated γ_s=max { γ_l| l=1,2 ..., N } i.e. the corresponding cross bearing of s-th track association occurrence matrix Point mark and radar target point trace space topological resemblance degree highest, then according to the non-zero entry of s-th track association occurrence matrix Element judges each radar and ESM track association situation under multi-target condition.

Claims (2)

1. under a kind of systematic error the radar based on topology information with ESM Data Association it is characterised in that comprising to walk as follows Suddenly：

Step one, acquisition cross bearing point：

The angular cross anchor point x of i-th radar target and j-th ESM target_ijAnd y_ijCoordinate

$\left.\begin{array}{c}{x}_{ij}\left(k\right)=\frac{\[x_A\cot \left({\mathrm{\θ}}_i^A\right(k\left)\right)-y_A\]-\[x_B\cot \left({\mathrm{\θ}}_j^B\right(k\left)\right)-y_B\]}{\cot \left({\mathrm{\θ}}_i^A\right(k\left)\right)-\cot \left({\mathrm{\θ}}_j^B\right(k\left)\right)}\\ y_{ij}\left(k\right)=\frac{\[y_A\tan \left({\mathrm{\θ}}_i^A\right(k\left)\right)-x_A\]-\[y_B\tan \left({\mathrm{\θ}}_j^B\right(k\left)\right)-x_B\]}{\tan \left({\mathrm{\θ}}_i^A\right(k\left)\right)-\tan \left({\mathrm{\θ}}_j^B\right(k\left)\right)}\end{array}\right\}---\left(1\right)$

Wherein, r_j(k) andIt is the range-azimuth measured value of k i-th radar target of moment respectively, (x_A,y_A) it is radar Coordinate, (x_B,y_B) for ESM position coordinate,Angle measurement for k j-th target of moment ESM；

Step 2, the local association based on range information：

Calculate the angular cross anchor point of i-th radar target and j-th ESM target away from radar apart from d_ij(k)

$d_{ij}\left(k\right)=\sqrt{{\left(x_{ij}\right(k)-x_A)}^2+{\left(y_{ij}\right(k)-y_A)}^2}---\left(2\right)$

Make D_ij(k)=d_ij(k)-r_i(k), because the measure error between radar angular and distance and the angle of ESM is separate, So D_ijK the variance of () is

Local based on range information thick association process rule is：WhenWhen, that is, think i-th radar Target is derived from same target with j-th ESM target, i.e. two flight path test associations；When being unsatisfactory for it is believed that i-th radar mesh Mark and j-th ESM target can not possibly be derived from same target；

Step 3, to remaining cross bearing dot-dash subassembly：

According to the principle with an ESM target association for each radar target, follow when A is split：From track association square Every a line of battle array A, selects unique 1 element, as track association occurrence matrix B_lIn this row and the unique non-zero of this row Element；

Wherein, A is after the local association based on range information, the radar obtaining and ESM potential track incidence matrix；B_l For track association occurrence matrix, obtained by splitting according to mentioned above principle,

Wherein, l=1,2 ..., N,For track association occurrence matrix B_lI-th row jth column element, is Boolean type variable；

Step 4, the overall gray relative based on topology information：

I-th target of l-th track association occurrence matrix with the space topological coefficient correlation sequence of i-th radar plot be

${\mathrm{\ξ}}_{li}^m\left(j\right)=\frac{\underset{l}{min}\underset{j}{min}|V_i^m\left(j\right)-U_{li}^m\left(j\right)|+\mathrm{\ρ}\underset{l}{\max }\underset{j}{\max }|V_i^m\left(j\right)-U_{li}^m\left(j\right)|}{|V_i^m\left(j\right)-U_{li}^m\left(j\right)|+\mathrm{\ρ}\underset{l}{\max }\underset{j}{\max }|V_i^m\left(j\right)-U_{li}^m\left(j\right)|}---\left(4\right)$

Wherein, U_liSpace topological for l-th track association i-th target of occurrence matrix is described as sequence, V_iFor the k moment i-th The topology information of individual radar plot is the alternate position spike sequence vector of itself and all neighbours, i=1,2 ... n, l=1,2 ..., N, j= 1,2 ..., n-1, m=1,2；

Because l-th track association occurrence matrix has n crosspoint, each crosspoint comprises n-1 neighbours, and each intersection Anchor point coordinate sequence all comprises two coefficient correlation phasesequence component of transverse and longitudinal axle, defines l-th track association occurrence matrix and thunder The gray relation grades reaching the similarity of the space topological of target are：

${\mathrm{\γ}}_l=\frac{1}{2(n-1)}\underset{i=1}{\overset{n}{\Σ}}\underset{j=1}{\overset{n-1}{\Σ}}\underset{m=1}{\overset{2}{\Σ}}{\mathrm{\ξ}}_{li}^m\left(j\right)---\left(5\right)$

It is calculated γ_s=max { γ_l| l=1,2 ..., N } i.e. the corresponding cross bearing point mark of s-th track association occurrence matrix With radar target point trace space topological resemblance degree highest, then sentence according to the nonzero element of s-th track association occurrence matrix Determine each radar under multi-target condition and ESM track association situation.

2. the radar based on topology information and ESM Data Association under a kind of systematic error according to claim 1, its It is characterised by, in step one, i-th radar is that radar target is passed according to angle with ESM target with j-th ESM target respectively Target after increasing order sequence.